This paper presents a study of artificial neural nets for use in stock index forecasting. The data from a major emerging market, Kuala Lumpur Stock Exchange, are applied as a case study. Based on the rescaled range analysis, a backpropagation neural network is used to capture the relationship between the technical indicators and the levels of the index in the market under study over time. Using different trading strategies, a significant paper profit can be achieved by purchasing the indexed stocks in the respective proportions. The results show that the neural network model can get better returns compared with conventional ARIMA models. The experiment also shows that useful predictions can be made without the use of extensive market data or knowledge. The paper, however, also discusses the problems associated with technical forecasting using neural networks, such as the choice of "time frames" and the "recency" problems.

/pdf/neural-networks-for-technical-analysis-a-study-on-klci-3qeezd92dx.pdf

Neural networks for technical analysis: a study on klci

Thin Trading and Stock Market Efficiency: the Case of the Kuala Lumpur Stock Exchange

Comparatively few studies have analyzed the social behavior of multinational corporations (MNCs) at a cross-national level. To address this gap in the literature, we propose a “transnational” model of corporate social responsibility (CSR) that permits identification of universal domains, yet incorporates the flexibility and adaptability demanded by international research. The model is tri-dimensional in that it juxtaposes: 1) Bartlett and Ghoshal’s (1998, 2000) typology of MNC strategies (multinational, global, “international,” and transnational); 2) the three conceptual domains of CSR (human rights, labor, and the environment) proposed by the UN Global Compact (Compact 2003); and 3) Zenisek’s (1979) description of three CSR perspectives (ideological, societal, and operational). The end result is a multidimensional typology that permits the organization and development of empirical CSR research in an international setting.

https://www.cambridge.org/core/services/aop-cambridge-core/content/view/S1052150X00007326

Transnational Corporate Social Responsibility: A Tri-Dimensional Approach to International CSR Research

This empirical study is conducted to test the weak-form market efficiency of the stock market returns of Pakistan, India, Sri Lanka, China, Korea, Hong Kong, Indonesia, Malaysia, Philippine, Singapore, Thailand, Taiwan, Japan and Australia. Monthly observations are taken for the period January 2004 to December 2009. Autocorrelation, Ljung-Box Q-statistic Test, Runs Test, Unit Root Test and the Variance Ratio are used to test the hypothesis that the stock market follows a random walk. Monthly returns are not normally distributed, because they are negatively skewed and leptokurtic. In aggregate we concluded that the monthly prices do not follows random walks in all the countries of the Asian-Pacific region. The investors can take the stream of benefits through arbitrage process from profitable opportunities across these markets.

Testing the Weak Form of Efficient Market Hypothesis: Empirical Evidence from Asia-Pacific Markets

Previous weak-form efficiency tests of the Kuala Lumpur and Singapore stock markets have mixed findings but mostly have suggested close conformity with a random walk. Using a disparate type of sample data over a longer and diverse time period, this paper seeks to replicate the earlier investigations of these two thin markets in order to verify whether the previous results may be sample specific. The distribution of daily stock returns is also examined to determine the validity of using statistical tests based on the normality assumption. Results tend predominately to confirm independence of serial stock returns and indicate distinctly nonnormal distributions.

Weak-form efficiency in the Kuala Lumpur and Singapore stock markets

* The stock price behavior of most developed countries has been known to follow the random walk process, eg, in Fama's (3) study on US stocks, Soinik (7), Theil and Leenders (8), and Dryden's (2) work on European stocks An interesting result emerges from the comparison of these studies: that although the random walk hypothesis would adequately depict the behavior of European stock markets, the deviation from a random walk is more apparent in European exchanges than in the US An accepted explanation takes into account the differences in the institutional characteristics of the two markets; such as, stringencies of disclosure requirements, control on inside trades, thinness of markets, and discontinuities of trading, etc (7) If the degree of "structure" or "organization" affects the degree of conformity to the random walk process, then we should expect countries, whose stock exchanges are not as "organized" as the European or US exchanges-eg, lesser degree of regulations or disclosure requirements-to exhibit greater deviation from a random price behavior Consequently, it would be interesting to see the results if similar tests were performed on stock prices from nonUS or European exchanges In the following, we will discuss results of serial correlation for five Far Eastern countries: Australia, Hong Kong, Japan, Philippines, and Singapore These results are then compared to those of the previous studies on US and European stocks

A Note on the Price Behavior of Far Eastern Stocks

I n recent years there has been an inaeasing awareness of the need to measure institutional portfolio performance. Most of the studies in this area have been associated with measuring the typical mutual funds, beginning with the studies by Jensen [l], and Treynor and Mazuy [2]. There are two problems associated with measuring the performance of these institutional portfolios that have been pointed out as a result of previous studies. First, as Levy [3] pointed out, the periods over which performance is measured should coincide with the planning horizons of the funds concerned. This task is very difficult at best. If the period selected is too long, the positive and negative effect during shorter time periods will be lost because of a cancelling out effect. The time periods used must then be flexible to cover the approximated planning horizons of the funds. The second problem of measuring institutional portfolio performance is trying to measure both the ability of the fund's management to choose whether securities are undervalued or overvalued and the timing of the market trends and reverses. For example, in the Treynor and Mazuy study [3], they measure the presence or absence of curvature in the characteristic line over the entire period (1953-1962) and do not attempt to measure the success of the fund managers in selecting overpriced or underpriced investments. The current study proposes a method to evaluate portfolio performance that reduces the severity of these two previously mentioned problems. This study deals with the performance of hedge funds. Publicly held hedge funds, investment companies that by their charter can buy on margin, sell short, hold warrants, 69 E

Randolph A. Pohlman

Papers

A Note on the Price Behavior of Far Eastern Stocks

Performance and Timing: A Test of Hedge Funds

Policies of multinational firms: A survey